HOW ARE SHORTS INFORMED? SHORT SELLERS, NEWS, AND … · 2020-07-07 · Kenan-Flagler Business...
Transcript of HOW ARE SHORTS INFORMED? SHORT SELLERS, NEWS, AND … · 2020-07-07 · Kenan-Flagler Business...
Electronic copy available at: http://ssrn.com/abstract=1535337
HOW ARE SHORTS INFORMED?
SHORT SELLERS, NEWS, AND INFORMATION PROCESSING*
Joseph E. Engelberg
Kenan-Flagler Business School, University of North Carolina
Adam V. Reed
Kenan-Flagler Business School, University of North Carolina
Matthew C. Ringgenberg
Kenan-Flagler Business School, University of North Carolina
FEBRUARY 22, 2010†
ABSTRACT
Combining a database of short sellers’ trading patterns with a database of news releases, we
show that short sellers’ trading advantage comes largely from their ability to analyze publicly
available information. Specifically, the prior finding that short sellers’ trades predict future
negative returns (e.g., Boehmer, Jones, and Zhang (2008) and Asquith, Pathak, and Ritter
(2005)) is more than twice as strong in the presence of news stories. Further, the most profitable
short sales do not appear to come from market makers, but from clients, and these client short
sales are particularly profitable in the presence of news. We also show that the ratio of short
sales to total volume is nearly constant around news periods, and when we do find differences
between the timing of short sellers’ trades and the overall market, we find that relative to other
types of trading there is a significant increase in short selling after news stories. Finally, short
sellers’ ability to predict returns appears to be concentrated in many of the news categories in
which short sellers trade relatively late; a finding consistent with the idea that short sellers’
advantage arises from their ability to process publicly available information.
*The authors thank Paul Tetlock for assistance with the Dow Jones news archive and we thank Dow Jones for
providing access to their news archive. We have benefited from comments from Greg Brown, Jennifer Conrad.
Wayne Ferson, Günter Strobl and Robert Whitelaw. We also thank seminar participants at the University of North
Carolina and the 2010 Utah Winter Finance Conference. This paper was previously titled “Buy on the Rumor,
[Short] Sell on the News: Short Sellers, News and Information Processing.” †Comments welcome. © 2010 Joseph E. Engelberg, Adam V. Reed, and Matthew C. Ringgenberg.
Electronic copy available at: http://ssrn.com/abstract=1535337
1
There is overwhelming evidence that short sellers are informed traders. In particular, a
number of empirical papers find that short selling predicts future returns (Asquith and Meulbroek
(1995), Senchack and Starks (1993), and Boehmer, Jones, and Zhang (2008)). Return
predictability, however, tells us little about how short sellers obtain an informational advantage
over other traders. In this paper we address this question by combining a database of public
news events with a database of all short sale trades, a unique combination that allows us to
comprehensively examine the relation between short selling and the release of public
information.
One aspect of the relation between short sales and news that has received a lot of
attention in the literature is timing. Short sellers have been shown to trade before public
information is released. For example, Karpoff and Lou (2009) show that short selling increases
before the initial public revelation of firms’ financial misrepresentation. Similarly, Christophe,
Ferri, and Angel (2004) find evidence of informed short selling in the five days before earnings
announcements. The financial crisis has also been linked to the timing of short sellers’ trades,
with the Securities and Exchange Commission suggesting that short sellers spread “false rumors”
in an effort to manipulate firms “uniquely vulnerable to panic.”1
To examine whether short sellers’ informational advantage is due to timing, we begin by
looking for evidence of abnormal short selling ahead of news events in the U.S. over the 2005 to
2007 period, a pattern that would be consistent with anticipation. We find no such pattern. In
fact, we find that the ratio of short sales to total volume is nearly constant around news events.
Further, when we do find differences between the timing of short sellers’ trades and the overall
1 “What the SEC Really Did on Short Selling,” by Chairman Christopher Cox, 24 July 2008, The Wall Street
Journal.
2
market, we observe that, relative to other types of trading, there is a significant increase in short
selling after the news event. This result indicates that, on average, short sellers trade on publicly
available information, that is, they do not uncover and trade on information before it becomes
public.
Given the finding that short sellers trade on publicly available information, we next
explore whether short sellers’ informational advantage is due to their superior ability to process
public information. Several papers find that abnormal short selling or high short interest
unconditionally predicts lower future returns (see, e.g., Asquith and Meulbroek (1995), Senchack
and Starks (1993), Boehmer, Jones, and Zhang (2008)). We find that abnormal short selling
does indeed lead to lower future returns, but that this effect is largely concentrated around news
events: short selling’s predicative effect on future returns is more than twice as strong in the
presence of news stories. Thus, a short seller’s most informative trades appear to be those in
response to newly released public news, which is consistent with short sellers being good
processors of information.
An alternative explanation for the above result may be that some buyers make systematic
mistakes around news events (Antweiler and Frank (2006)), and that these buyers’ mistakes are
reflected in market makers’ offsetting short sales. To determine whether short sellers’ trades are
due to superior information processing or to offsetting positions, we exploit a unique feature of
the short selling data, namely, exempt versus non-exempt trade marking, to distinguish market
makers from non-market makers (or clients). We find that clients’ trades are particularly well
informed, and that these trades are much more profitable in the presence of news events. In
contrast, market makers’ trades are not particularly well informed, and there is no differential
impact in the presence of news. Overall, we conclude that the most informed short sales are
3
from clients, and that these shorts are particularly well informed in the presence of recent news.
This evidence lends support to the view that short sellers’ information advantage is due to their
superior information processing ability.
In our next set of tests we identify which types of information are associated with short
sellers’ advantage. To do so, we use the news subject classification in the Dow Jones archive to
sort stories into various categories ranging from analyst comments to earnings announcements to
new debt issues. We find that short sellers’ most informative trades are concentrated in five
categories: Corporate Restructurings, Earnings, Earnings Projections, New Products &
Services, and Stock Ownership. Further, many of these categories correspond to the categories in
which short sellers’ trades are measurably later than other investors’ trades, which lends
additional support to the idea that short sellers’ advantage stems from superior ability to process
publicly available information rather than an ability to uncover information before it becomes
publicly available.
Finally, we examine the economic significance of traders’ ability to trade on news by
implementing a portfolio approach. Recognizing that the presence of news is likely correlated
with firm characteristics and that some categories of news may be more relevant for some firms
than for others, we conduct an experiment in which each firm’s response to a news event is
matched by a similar firm’s response on the same day. We find that across all news categories,
short sellers’ advantage in predicting returns is concentrated in firms with news.
The findings of this paper shed light on the broader debate about the informational effects
of news announcements. While several papers argue that public news announcements reduce
information asymmetry (e.g., Korajczyk, Lucas, and McDonald (1991), Kacperczyk and Seru
4
(2007)), another thread of the literature argues that public information events present profitable
trading opportunities for skilled information processors (e.g., Engelberg (2008), Demers and
Vega (2008)), in effect increasing the asymmetry of value-relevant trading signals. Tetlock
(2009) weighs in on this question by modeling liquidity shocks and news and finds empirical
evidence suggesting that public information plays a key role in informing a subset of investors.
By focusing specifically on short sellers, this paper has a unique ability to shed light on the trade-
level evidence in this debate for two reasons: first, a number of papers show that short sellers are
informed (e.g., Asquith and Meulbroek (1995), Diamond and Verrecchia (1987)); second, short
sellers’ trades are among the few classes of trades that are uniquely identified. Our finding that
short sellers’ trades are more than twice as profitable in the presence of recent news is strong
evidence in favor of the idea that news presents profitable trading opportunities for skilled
information processors.
The remainder of this paper proceeds as follows. Section I discusses related literature.
Section II describes the databases used in this study. Section III presents our analyses and
findings. Finally, Section IV concludes.
I. Related Literature
The ideas in this paper relate to three distinct branches of the existing literature. First,
this paper relates to an extensive literature on the behavior of short sellers relative to other
traders. Second, our paper contributes to a growing literature on how market participants
respond to public news. Finally, this paper sheds light on an emerging debate on whether news
increases or decreases information asymmetry. In this section, we first discuss prior papers that
5
connect news to short selling. We then provide an overview of the relevant literature in each of
these three branches.
Several extant papers look at short selling behavior in the context of a single type of
corporate news event. As such, these studies shed light on a subset of this paper’s sample of
news events. Karpoff and Lou (2009), for example, examine short sellers’ positions in firms that
are investigated for financial misconduct; they find that short sellers generally anticipate public
announcements of investigations. Focusing on short sellers’ trades around earnings
announcements, Christophe, Ferri, and Angel (2004) find that short sellers do not tend to trade
before earnings announcements. Similarly, Daske, Richardson, and Tuma (2005) look at short
selling around earnings announcements and management forecast announcements and find no
evidence that short sale transactions concentrate prior to bad news events. Nagel (2005) looks at
the cash flow news implied by a vector auto regression and finds an asymmetric effect on
returns, indicating that short sellers help incorporate news into prices when short selling is not
constrained. Finally, Edwards and Hanley (2008) examine short selling around IPOs, a
newsworthy corporate event, and find evidence that casts doubts on short sale constraints as an
explanation for IPO pricing anomalies.
In contrast to the above papers, which identify patterns in short selling around specific
corporate new events, the current paper aims to uncover patterns in short sellers’ trades around
all types of corporate news events. In doing so, we try to understand short sellers’ behavior more
generally.
A. Short Sellers’ Trading Patterns
6
Several papers compare the trades of short sellers to the trades of other market
participants. There are several dimensions over which trades can be compared. Much of the
recent literature focuses on the profitability of trades, which roughly speaking, can be measured
as the performance of a stock’s price after the short sale. One of the most widely cited results in
this vein of the literature is found in Asquith and Muelbroek (1995), who show high short
interest precedes negative future returns, consistent with informed trading. Similarly, Asquith,
Pathak, and Ritter (2005) show that when short selling is constrained and there are relatively
diverse opinions, in some cases abnormally high short interest precedes negative future returns.
Using transaction data at a higher frequency, Boehmer, Jones, and Zhang (2008) find that
heavily shorted stocks significantly underperform lightly shorted stocks, and Diether, Lee, and
Werner (2008) show that not only do prices follow short selling, but short selling also follows
prices, that is, short sellers tend to short after price run ups. These results further indicate that
short sellers may have an informational advantage.2
In sum, the prior work above establishes that the performance of short sellers’ trades
indicates that short sellers’ trades are informed. Our paper contributes to this literature by asking
how short sellers come to enjoy an informational advantage in the first place.
B. Public News
2 A closely related dimension of research is whether short sellers’ trades reveal information to other market
participants. In other words, are short sellers’ trades news worthy in and of themselves? Senchack and Starks
(1993) show that abnormally large short interest announcements have small but significant negative returns.
Similarly, Aitken, et al. (1998) show that short sales are followed by price declines within 15 minutes on the
Australian Stock Exchange.
7
While a large literature examines volume and return phenomena around specific news
events (e.g., earnings announcements, mergers, and dividend initiations and omissions), a more
recent literature considers such phenomena around any corporate news event. Categorizing all
Wall Street Journal stories, between 1973 and 2001, Antweiler and Frank (2006) find that return
responses vary widely across news categories, although they find evidence of overreaction
(return reversal) on average. Also using a database of all news events, Tetlock (2008) finds
evidence of even stronger return reversal following repeated news events consistent with the idea
that investors overreact to “stale” news stories. Furthermore, using comprehensive news
databases, several studies examine whether well-known asset pricing anomalies are related to
news. Chan (2003) considers the momentum anomaly among stocks with and without recent
news and finds evidence of price momentum only among news stocks. Similarly, Vega (2006)
finds more earnings momentum among stocks with high differences of opinion on news days.
More recently, researchers have asked whether the content of news stories contains
value-relevant information. Tetlock, Saar-Tsechansky, and Macskassy (2008) and Engelberg
(2008) show that, indeed, the qualitative content of the information contained in news stories can
predict both earnings surprises and short-term returns. These findings support the idea that there
is value-relevant or “soft” information in news stories that is not immediately impounded into
prices.
To summarize, this literature highlights the importance of looking at more than one news
category in assessing short sellers’ behavior, and shows that the information content of news
leaves room for traders with different abilities to process the information to arrive at different
conclusions about the value relevance of the news. Our work builds on these findings by
analyzing the universe of corporate news events in the U.S. over our sample period, and by
8
asking whether, in our sample, information processing ability plays a role in the performance of
short sellers’ trades.
C. Public News and Informed Trading
There are two views regarding the relation between public news, such as the articles
available in the Dow Jones archive, and the trading of skilled investors. Under the first view,
public information does not provide traders with an information advantage, that is, managers
who rely on public information (rather than generate private information) are low-skilled.
Consistent with this view, Kacperczyk and Seru (2007) estimate managers’ reliance on public
information (RPI) as the R-squared of a regression of percentage changes in fund managers’
portfolio holdings on changes in analysts’ past recommendations and find that fund managers
with low RPIs (low reliance on public information) perform better than fund managers with high
RPIs (high reliance on public information).
Under the alternative view, the public release of information presents trading
opportunities for skilled processors of information, that is, when news is released, traders with
superior information processing skills can convert this news into valuable information upon
which to trade. Earnings announcements, for example, are often accompanied by lengthy
documents and conference calls that are scrutinized by information processors. Those traders
who show exceptional skill in converting such data into value-relevant information are rewarded
with superior returns on event-driven trades. Evidence consistent with this view comes from
studies that attempt to look at the textual content of news and firm announcements. Specifically,
Tetlock et al. (2008), Engelberg (2008), Demers and Vega (2008), and Feldman et al. (2009) all
9
show that the content of corporate news predicts returns, which is consistent with the view that
information processing skills can generate superior returns.
Our paper sheds light on the above debate by finding additional evidence in support of
the second view, that is, by showing that trades occurring after the release of news stories can be
more profitable than trades in non-news periods.
II. Data
The data used in this study come from the intersection of two databases. The first
database contains information on short sales while the second contains news articles from the
Dow Jones News Service. Below we describe the two databases in turn.
A. Short Sales
Information on short sales transactions comes from the NYSE TAQ Regulation SHO
database. Regulation SHO was adopted by the SEC in June of 2004 to establish new rules
governing short sales in equity transactions and to evaluate the effectiveness of price test
restrictions on short sales. As one consequence of regulation SHO, transaction-level short sales
data were publicly disclosed for the period January 3, 2005 through July 6, 2007. The NYSE
TAQ regulation SHO database therefore contains data for all short transactions that were
reported on the NYSE during this period. Specifically, the database contains the stock ticker, the
date and time of the transaction, the number of shares traded, the execution price, and an
indicator that denotes whether the transaction was exempt from price test rules. One of the
10
reasons a short sale transaction could be classified as exempt is that it was made by market
makers engaged in bona fide market making activity. The exempt indicator has thus been used
to separate trading by market makers from trading by non-market makers (e.g., Evans et al.
(2009), Christope, Ferri, and Angel (2004), Boehmer, Jones, and Zhang (2008), Chakrabarty and
Shkilko (2008)).3 However, when regulation SHO was implemented, a group of randomly
selected stocks was selected to be part of a pilot study for which the exempt/non-exempt
classification was no longer required. We exclude these pilot firms when using the exempt
indicator variable in our analyses.4
For the purposes of our analysis, we aggregate the transaction data at the daily level, and
we use the TAQ master files to add CUSIPs to the database. We then use the CRSP Daily Stock
Event file to add PERMNOs to the database. Finally, we add returns, total volume, and shares
outstanding information that we obtain from CRSP.
B. Dow Jones Archive
To compile our sample of news events, we use the Dow Jones archive as in Tetlock
(2009). This archive contains all Dow Jones News Service stories and Wall Street Journal
stories over our 2005 to 2007 sample period.
The Dow Jones database also contains subject codes that identify the information content
of each news article; for example, there is a code to indicate that an article contains information
3 For example, NASD NTM 06-53 notes that “Rule 5100(c)(1) provides an exception to the bid test for short sales
by a market maker registered in the security in connection with bona fide market making activity.” 4 Details regarding the regulation SHO pilot study, including a list of firms involved, are available on the SEC
website: http://www.sec.gov/rules/other/34-50104.htm. Our results are robust to the inclusion of the regulation SHO
pilot firms.
11
about insider stocks sales. We adopt Dow Jones’ subject categorizations. Starting with the
database described in Tetlock (2009), we have 71 news categories. However, many of these
subject codes are general codes that do not provide valuable information about the content of a
news article. For example, nearly every article in the database has the code Company News
assigned to it, in addition to a more specific news code. We remove these general codes from
our analysis to obtain a final list of subject codes that contains 39 different news categories.5
The resulting news database contains a unique firm identifier, subject codes, a dummy
variable that takes the value of one if a story was released in multiple pieces over the news day,
and two sentiment score variables that indicate whether a story contains negative words in the
headline and body of the text. The first sentiment variable is constructed using the Harvard-IV-4
dictionary as in Tetlock (2007) and Engelberg (2008) while the second sentiment measure uses
the negative word list developed by Loughran and McDonald (2009). In both cases, we
construct the sentiment score as the sum of the number of negative words in an article’s headline
and body divided by the sum of the total number of words in the headline and body.
We use the unique firm identifier to match the news data to the short sales database. The
resulting database has 1,888,868 observations over the period January 3, 2005 to July 6, 2007.
Table I contains summary statistics for the combined database. The mean number of articles per
firm-day is 1.10. However, there is substantial cross-sectional variation in this number, and
larger firms typically have more news articles on a given day. Certain news categories also
appear much more often than others. For example, the category High Yield Issuers appears
173,357 times in the database while the category 10K appears only 1,320 times. To address the
5Specifically, after computing the correlations between subject codes, we exclude subject codes if their correlation
with a more specific news category exceeds 80%. We also drop news categories that are associated with fewer than
1,000 news events over the entire sample (see Table I for the frequency of each news event in our sample).
12
potential issue of news clustering, when we conduct category specific analyses we remove
stories that are within 30 days of another story in the same category.
III. Analyses and Results
In this section we explore how short sellers differ from other traders. We begin by asking
whether short sellers respond to news before other market participants. We find that short sellers
tend to trade at the same time as other traders, and when they do not, they trade after other
traders. These results suggest that short sellers do not anticipate news. Next, we ask whether
short sellers’ trades are more profitable than other trades, consistent with a superior ability to
process news. We find consistent evidence. In a third set of tests we analyze which types of
information are associated with short sellers’ profitability. Finally, we conduct a matched
sample portfolio approach to shed additional light on the economic impact of news-based short
sales strategies.
A. Do Short Sellers Anticipate News?
One way in which short sellers may differ from other traders is in the timing of their
trades. There is some evidence that short sellers anticipate bad news announcements (e.g.,
Angel, Ferri, and Christophe (2004) and Karpoff and Lou (2009)). However, these findings
correspond to specific types of corporate events. Here we seek to shed light on short sellers’
timing behavior around all types of news events in our sample period.
13
To determine the extent of short sales timing around news events, in Figure 1 Panel A we
plot daily short sales volume (solid line), total volume (dashed line), and the ratio between the
two (dotted line) in calendar time around our universe of news events. The basic result is readily
apparent: short sellers trade when other traders do. Clearly, all traders respond to news, as there
is a significant increase in volume on the news event day and on surrounding days. However,
the ratio of short sales to total volume is nearly constant over the news period, with no significant
change in the ratio around news events. This result suggests that short sellers do not uncover and
trade on information before it becomes publicly available.
Of course, in line with the prior research above, it may be the case that short sellers
respond more to certain types of news, particularly bad news. Thus, in Panels B and C of Figure
1, we focus only on negative news events, where negative news events are defined using the
Harvard-IV-4 dictionary (see Section II.B) and the Loughran and McDonald (2009) negative
word list. The results are largely unchanged, indicating that the timing of short sellers’ response
to news does not depend on whether the news is bad.
Next, we assess whether the timing of short sellers’ trades varies by news category.
Table II presents results from a regression of short sales volume on a set of indicator variables
representing each of the news categories. Specifically, we regress daily aggregate short volume
on indicator variables that take the value one if there is a news story in a particular news category
on a given day and zero otherwise. To control for the short sellers’ response to past returns (e.g.,
Diether, Lee, and Werner (2008)), we include two lags of daily returns. The results indicate that
for the majority of news categories, short sellers respond at the same time as other traders. More
specifically, for a given news event, when we compare the coefficient estimates for regressions
on short selling after the news event to estimates for regressions on short selling before the news
14
event, we find that the coefficient estimates are largely the same. However, there are a number
of interesting exceptions. For both types of earnings news stories, Earnings and Earnings
Projections, there is more short selling after the news event than before the news event, a result
largely consistent with the findings of Angel, Ferri, and Christophe (2004). The statistically
significant estimate of 0.0049 in the t+2 specification for Earnings indicates that there is a 0.49%
increase in short selling as a percentage of total volume two days after news of this type is
reported. This late response is also apparent in news stories about joint ventures and product
distribution. In contrast, stories about leveraged buyouts show the opposite pattern. The
estimate of 0.0145 in the t-1 specification indicates that the short selling ratio increases 1.45% on
the day before news stories about leveraged buyouts, and the statistically significant coefficient
estimate on After Minus Before indicates that, relative to the two-day period before the news
event, the short selling ratio decreases 2.33% in the two-day period after the news event.
Table III presents results from a similar setup, but where the dependent variable is raw
daily short sales volume rather than short sales volume scaled by total volume. The change in
raw sales volume can be interpreted as a direct measure of the average increase or decrease in the
number of shares traded in response to news events. Our findings are qualitatively unchanged.
For instance, in the most extreme case, news days that contain Money Market News are
associated with a statistically significant increase of 291,721 additional shares sold short.
Despite the natural interpretation of these results, however, they are not as meaningful as the
scaled results discussed above. The reason is that over our sample period there is a strong
market-wide trend towards greater short volume, and thus in this analysis After Minus Before is
15
statistically significant in approximately half of the news categories.6 To summarize, in this
subsection we show that short sellers generally trade at the same time as other traders, and in
those instances in which they show different timing, short sellers tend to trade after other traders.
This suggests that short sellers’ previously documented information advantage (e.g., Boehmer,
Jones, and Zhang (2008) and Asquith, Pathak, and Ritter (2005)) does not stem from an ability to
anticipate news.
B. Do Short Sellers Have Superior Information Processing Ability?
Given our finding above, in this subsection we ask whether short sellers’ informational
advantage derives from an alternative source, namely, a superior ability to process the
information contained in publicly available news.
To answer this question as directly as possible, we begin by replicating Table IV of
Boehmer, Jones, and Zhang (2008), shown in our Table IV below. Specifically, we compute 20-
day rolling returns (i.e., t+1 to t+21) from January 3, 2005 through July 6, 2007 and we regress
these returns on the Short Volume Ratio on day t, which is defined as daily short volume divided
by total volume. The Boehmer, Jones, and Zhang (2008) result comes through strongly in these
results: in each of the specifications, Short Volume Ratio is negative and statistically significant,
indicating that when there is an increase in short sales, future prices decrease. However, given
our previous results, we might expect this pattern to be stronger among firms for which news is
released when short volume is high. To test for this effect, we include the indicator variable
6 Mean daily short volume increases from 120,910 shares in 2005 to 150,681 shares in 2007, and the increase is
statistically significant. This steady increase in short volume through time offers an explanation for the fact that
abnormal volume in Figure 1 is generally slightly above 1.
16
News Event, which takes the value one if there is news on day t and zero otherwise. We also
include contemporaneous returns as a control for the information content in the news, and we
include two days of lagged returns to control for the tendency of short sellers to trade following
recent price increases as documented by Diether, Lee, and Werner (2008). Note that because
news coverage is correlated with firm characteristics such as size and institutional ownership
(e.g., Chan (2003), Vega (2006), Engelberg (2008), and Fang and Peress (2009)), our empirical
design is meant to estimate the effect of news within firms rather than across firms. We thus
follow Skoulakis (2005) and apply the Fama-Macbeth approach to firms: we first run a time-
series regression for each firm; we then take the average of the coefficients and use the standard
deviation to estimate standard errors.
The results, shown in Table IV, provide strong evidence on the informational advantage
of short sellers. Specifically, the coefficient estimate of -0.0053 on Short-News Interaction in
Model 5 is negative and statistically significant, indicating that among stocks with high short
volume, those with news have significantly more negative future returns than those without
news.7 Even after controlling for the contemporaneous effect of returns, the coefficient on Short
Volume Ratio is still negative at the 5% level, in other words, the Boehmer, Jones, and Zhang
(2008) finding that short volume leads negative returns continues to hold. Our findings thus
provide new insight into the source of short sellers’ informational advantage. In particular, we
7 In unreported results, we add the number of negative words (a measure of the sentiment in the article) as a control
variable. This variable does not change the general magnitude or statistical significance of the results, indicating
that the findings are not driven by either very good or very bad news events.
17
find that the previously documented relation between short volume and returns is more than
twice as strong for those stocks that have a public news event.8
A drawback to using total short sales volume as a measure of short selling is that some
short sales are generated as a result of market making – to the extent that some buyers make
systematic mistakes, the corresponding short sales are simply offsetting positions, not informed
trades. Thus, with the aggregate measure of short sales volume used in Table IV, we cannot
distinguish the effect of short sales that arise in response to counterparty purchases from the
effect of shorts that arise for the purpose of gaining negative exposure. This raises the question
of whether our results in Table IV can be attributed to informed trading. To address this
concern, we take advantage of a unique feature of the data, namely, the exempt versus non-
exempt classification of trades. This classification allows us to separate shorts into market
making and non-market making (i.e., client) trades.9
Tables V and VI report the results for non-exempt and exempt trades, respectively. In
Table V the statistically significant coefficient estimate of −0.0075 in Model (5) indicates that
high short volume is a significant predictor of low future returns. Moreover, the magnitude on
short volume is 44% larger than the corresponding coefficient for total short sales volume in
Table IV, which suggests that the ability of short sales to predict future returns is particularly
8 The Boehmer, Jones, and Zhang (2008) result can be thought of as a high-frequency analog of the results in
Asquith and Muelbroek (1996) and Asquith, Pathak, and Ritter (2005). This second set of papers measures short
trading with short interest instead of short volume, and they use future returns that are measured over longer periods.
Although we would like to examine the relation between news and short sellers’ advantage in the context of these
short interest-based findings, there is an econometric challenge in making a direct comparison. Specifically, news in
our database is marked with daily time stamps, so either we would have to aggregate news to match the monthly
frequency of short interest or we would have to throw out much of our news data. It is not clear how a reduction in
the frequency of the news variable would change expectations about the short positions. 9 Anecdotal evidence suggests that the exemption is sometimes abused, but only in one direction: trades may be
inappropriately marked as exempt when they are not. Since the exemption removes potential restrictions, it is
unlikely that exempt trades would ever be inappropriately marked as non-exempt. In other words, exempt trades
may include client trades, but non-exempt trades are unlikely to include market maker trades.
18
strong for non-market-making trades. We also see that the short-news interaction estimate of
−0.0056 is significantly negative, indicating that non-market makers’ shorts are 75% (i.e.,
−0.0056/−0.0075) more profitable in the presence of news events than at other times. In
contrast, the results in Table VI indicate that market makers’ trades are not particularly well
informed: the short volume ratio loads as a positive predictor of price, indicating that market
makers’ shorts are actually associated with positive future returns on average.10
Further, there is
no differential effect of market makers’ trading in the presence of news.
Overall, the evidence in this subsection suggests that the most informed short sales are
made for the purpose of gaining negative exposure, and that these trades are particularly well
informed in the presence of recent news events.
C. Price Responses by News Category
In an extension to the above analysis, in this subsection we ask whether short sellers’
information processing ability is uniformly strong across news categories. To get at this
question, we repeat the analysis in Table IV separately for each news category. Specifically, for
each news category we run a regression in which the dependent variable is the compound return
from the first to the twentieth trading day after the news event, and the main independent
variable is short vol / market vol, which is the amount of short selling relative to total volume on
the day of the news event. Since the type of news (good or bad) may have some effect on future
10
Even though the magnitudes of the coefficients vary between Table V and Table VI, the economic impact is of the
same order of magnitude. Specifically, a one standard deviation increase in short volume among non-exempt trades
leads to a 0.200% decrease in future returns, while a one standard deviation increase in short volume among non-
exempt trades is associated with a 0.154% increase in future returns.
19
returns (e.g., Bernard & Thomas (1989)), we attempt to control for whether the news is good or
bad by including the event-day return on the right-hand side of the specification.
The results, shown in Table VII, indicate that short sellers have some ability to identify
trades that are likely to be profitable around certain news events. Specifically, we find that the
coefficient estimates on short vol / market vol are significantly negative for 12 of the 39 news
categories; of these, five are statistically significant at the 1% level (Corporate Restructurings,
Earnings, Earnings Projections, New Products & Services, and Stock Ownership), and nine are
significant at the 5% level (Corporate Restructurings, Divestitures or Asset Sales, Earnings,
Earnings Projections, Initial Public Offerings, Management Issues, New Products & Services,
Research and Development, and Stock Ownership). As a further test of statistical significance,
we conduct a Fisher test of combined probability to determine whether the cross-sectional
distribution of the p-values from each regression differs significantly from a uniform zero-one
distribution. The Fisher test rejects this null at the 1% level of significance across all news
categories, suggesting that the coefficient on short volume is statistically different from zero for
the cross-section. We also find that many of the categories that are statistically significant are
the same categories identified in Subsection A as the categories in which short sellers’ trades are
measurably later than other investors’ trades (e.g., Earnings, and Earnings Projections).
Taken together, these results indicate that when short selling predicts future returns, short
sellers appear to be making profitable trades. This evidence lends further support to the idea that
short sellers’ informational advantage stems from superior ability to process publicly available
information.
20
D. Matched Sample Portfolio Approach
So far we provide evidence that short selling is more informative on news event days. In
this subsection we shed light on the economic impact of news-based short selling strategies using
a portfolio approach. This approach recognizes that the presence of news is likely correlated
with firm characteristics and that certain categories of news may be more relevant for some firms
than for others. In other words, since news is strongly related to several firm characteristics, we
cannot simply sort on news. Moreover, news coverage is highly persistent: firms that have many
news articles in the Dow Jones archive in one year are likely to have many articles in following
years. Thus, in order to conclude that news, rather than a particular firm characteristic, is driving
the differential returns we observe we need to compare two firms that are identical apart from the
fact that one firm has a news event while the other firm does not. We do this using a matched
sample portfolio approach.
Our approach is based on forming portfolios of stocks around news events. Because
previous research indicates that firm characteristics may affect future returns, we implement a
control sample methodology to control for these previously documented effects. Specifically, for
every stock with a news event, we identify a control stock that is the closest match in the
following four dimensions: bid-ask-spread, institutional ownership, market capitalization, and
number of news events over the previous month. We match by selecting the stock that
minimizes the sum of the rank differences in each of these categories. Furthermore, to eliminate
potentially contaminating competitive effects (e.g., Slovin, Sushka, and Bendeck (1991), Chen,
21
Ho, and Ik (2005), and Hsu, Reed, and Rocholl (2009)), we require that control firms and sample
firms be members of different Fama-French 48 industries.11
The analysis yields results for each of the 39 news categories. Figure 2 presents the
results for three categories as examples. In the Dividends category, we see that among firms
with dividend news, firms with high short volume have significantly lower returns than firms
with low short volume. This difference is approximately 4.39% at the one-year point. In
contrast, the control sample shows similar returns across high short volume stocks and low short
volume stocks.
Table VIII summarizes the detailed results of this analysis. The economic significance of
news-based short sales becomes apparent when we compare differences in portfolio returns. For
example, if an investor were to sell a portfolio of stocks with high short selling and buy a
portfolio of stocks with low short selling on the day that Product Distribution news is released,
that investor would earn an excess annual return of 6.50%. The same strategy for a matched
portfolio of no-news stocks would return 5.74% over the period, yielding a difference of
12.24% annually between the two strategies. In fact, this strategy yields positive excess returns
in 34 out of our 39 news categories, with some news categories yielding annualized excess
returns of over 10%. The statistically significant 2.89% return for the mean excess return
difference indicates that not only do short sellers have a significant advantage over other traders,
but their advantage comes largely from their ability to process and trade on news events.
To summarize, this analysis shows that the inverse relation between short volume and
future returns is strongest around news events, whereas during non-news events this relation may
11
In unreported results, we use the Fama-French 12 industry classifications instead of the Fama-French 48
classifications. The results are not qualitatively different.
22
be insignificant or even go in the other direction. These results lend additional support to our
main finding that the previously documented informational advantage of short sellers is driven in
large part by short sellers’ superior ability to process information contained in publicly available
news.
IV. Conclusion
Previous research documents that short sellers are informed traders (e.g., Boehmer, Jones,
and Zhang (2008) and Asquith, Pathak, and Ritter (2005)). Yet we know little about the source
of short sellers’ informational advantage. This paper seeks to fill this gap by investigating the
following questions: To what extent are short sellers able to anticipate news events? Are short
sellers better able to process and react to news? And, are short sellers’ trades particularly
profitable around specific categories of news? To address these questions, we combine a
database of all public news events in the U.S. with a database of short sale trades over the same
sample period.
We find that, in general, short sellers trade at the same time as other traders. Specifically,
the ratio of short sales to total volume is nearly constant over news periods, with no significant
change in the ratio around news events. However, we do find some differences between the
timing of short sellers’ trades and the overall market: for news stories about analysts’ comments
and ratings, earnings, earnings projections, joint ventures, and product distribution, there is a
significant increase in short selling after the news story. This finding suggests that like other
traders, short sellers trade on publicly available information, and hence their informational
advantage is not due to an ability to uncover or anticipate information before it becomes public.
23
Given the result that short sellers’ advantage is not due to timing, we next ask if it could
be due to superior ability to process the information available in public news stories. We find
supportive evidence. In particular, we find that across all types of news, short selling predicts
future returns: even after controlling for the unconditional relation between short selling and
news (e.g., Boehmer, Jones, and Zhang (2008)), short selling’s predicative effect on future
returns is more than twice as strong in the presence of news. This result is not a reflection of
persistent mistakes by buyers, that is, the most informed short sales are not from market makers
but from clients, and these client shorts are particularly well informed in the presence of news.
We also find that this predicative effect is strongest for nine categories of news (Corporate
Restructurings, Divestitures or Asset Sales, Earnings, Earnings Projections, Initial Public
Offerings, Management Issues, New Products & Services, Research and Development, and Stock
Ownership), and that many of these categories are the same categories for which short sellers’
timing follows the overall market. Finally, recognizing that the presence of news is likely
correlated with firm size and that certain categories of news may be more relevant for some firms
than for others, we conduct an experiment in which each firm’s response to a news event is
matched by a control firm’s response on the same day. We find that across all news categories,
short sellers’ advantage in predicting returns is concentrated in firms with news.
In sum, we show that, on average, short sellers’ advantage is not due to an ability to
influence the public’s perception of value, as recently suggested by the Securities and Exchange
Commission.12
Rather, we find that short sellers generally trade when other traders do, and to
the extent that the timing of their trades differs, short sellers actually trade after other traders.
We further find that short sellers’ ability to predict future negative returns is concentrated around
12
Short sellers were accused of “distort and short” schemes in “What the SEC Really Did on Short Selling” by
Chairman Christopher Cox, 24 July 2008, The Wall Street Journal.
24
news events. Thus, by connecting short sellers’ trading patterns with news releases, we show
that short sellers’ trading advantage derives primarily from their superior ability to analyze
publicly available information.
The findings of this paper shed light on the broader debate about the informational effects
of news announcements. While several papers argue that public news announcements reduce
information asymmetry (e.g., Korajczyk, Lucas, and McDonald (1991), Kacperczyk and Seru
(2007)), others have recognized that public news events lead to differential interpretations by
traders (Kandel and Pearson (1995)) based on the skill of those traders. Rubenstein (1993) puts
it succinctly: “In real life, differences in consumer behavior are often attributed to varying
intelligence and ability to process information. Agents reading the same morning newspapers
with the same stock price lists will interpret the information differently.” This view explains not
only why volume is high around news events (Kandel and Pearson (1995)) but also why some
papers find return predictability from “soft” information in news announcements (e.g., Engelberg
(2008), Demers and Vega (2008)). Specifically, public information events present profitable
trading opportunities for skilled information processors. Tetlock (2009) weighs in on this
question by modeling liquidity shocks and news and finds empirical evidence suggesting that
public information plays a key role in informing a subset of investors. By focusing specifically
on short sellers, this paper has a unique ability to shed light on the trade-level evidence in this
debate for two reasons: first, a number of papers show that short sellers are informed (e.g.,
Asquith and Meulbroek (1995), Diamond and Verrecchia (1987)); second, short sellers’ trades
are among the few classes of trades that are uniquely identified. Our finding that short sellers’
trades are more than twice as profitable in the presence of recent news is strong evidence in favor
of the idea that news presents profitable trading opportunities for skilled information processors.
25
REFERENCES
Antweiler, W. and M. Frank, 2006, Do U.S. stock markets typically overreact to corporate news
stories?, Working Paper, University of British Columbia.
Asquith, P., and L. Meulbroek, 1995, An empirical investigation of short interest, Unpublished
Working Paper, M.I.T.
Asquith, P., P. Pathak, and J. Ritter, 2005, Short Interest, Institutional Ownership, and Stock
Returns, Journal of Financial Economics 78, 243-276.
Aitken, M., A. Frino, M. McCorry, and P. Swan, 1998, Short sales are almost instantaneously
bad news: Evidence from the Australian Stock Exchange, Journal of Finance 53, 2205-
2223.
Bernard, V. and J. Thomas, 1989, Post-earnings-announcement drift: delayed price response or
risk premium?, Journal of Accounting Research 27, 1-36.
Boehmer, E., C. Jones and X. Zhang, 2008, Which Shorts are Informed?, Journal of Finance 63,
491-527.
Chakrabarty, Bidisha, and Andriy Shkilko, 2008, Information Leakages in Financial Markets:
Evidence from Shorting around Insider Sales, Working Paper.
Chan, W., 2003, Stock price reaction to news and no-news: drift and reversal after headlines,
Journal of Financial Economics 70, 223-260.
Chen, S., K. W. Ho, and K. H. Ik, 2005, The Wealth Effect of New Product Introductions on
Industry Rivals, Journal of Business 78, 969-996.
Christophe, S., M. Ferri, and J. Angel, 2004, Short-Selling Prior to Earnings Announcements,
Journal of Finance 59, 1845-1875.
Daske, H. S. Richardson, and A. Tuma, 2005, Do Short Sale Transactions Precede Bad News
Events?, Working Paper.
Diamond, D., and R. Verrecchia, 1987, “Constraints on short-selling and asset price adjustment
to private information,” Journal of Financial Economics, 18, 277–311.
Diether, K., K. Lee, and I. Werner, 2008, Short-sale Strategies and Return Predictability, Review
of Financial Studies 22, 575-607.
Edwards, A. and K. Hanley, 2008, Short Selling in Initial Public Offerings, Working Paper.
Engelberg, J., 2008, Costly Information Processing: Evidence from Earnings Announcements,
Working Paper, University of North Carolina.
26
Evans, Richard, Chris Geczy, David Musto and Adam Reed, 2009, “Failure is an Option:
Impediments to Short-Selling and Options Prices”, The Review of Financial Studies
22(5), 2009.
Fang, L. and J. Peress, 2009, Media Coverage and the Cross-Section of Stock Returns,
Forthcoming in the Journal of Finance.
Feldman, Ronen, Suresh Govindaraj, Joshua Livnat, and Benjamin Segal, 2008, The incremental
information content of tone change in management discussion and analysis, Working
paper, INSEAD.
Fox, Merritt B., Lawrence Glosten, and Paul Tetlock, 2009, Short Selling and the News: A
Preliminary Report on an Empirical Study, Working Paper.
Gervais, S., R. Kaniel, and D. Mingelgrin, 2001, The High-Volume Return Premium, Journal of
Finance 56, 877-919.
Hsu, H., A. Reed, and J. Rocholl, 2009, The new game in town: competitive effects of IPOs,
Forthcoming in the Journal of Finance.
Kacperczyk, M. and A. Seru, 2007, Fund Manager Use of Public Information: New Evidence on
Managerial Skills, Journal of Finance 62, 485-528.
Kandel, Eugene, and Neil D. Pearson, 1995, Differential interpretation of public signals andtrade
in speculative markets, Journal of Political Economy 103, 831-872.
Karpoff, J. and X. Lou, 2009, Short sellers and financial misconduct, Working Paper.
Loughran, Tim and Bill McDonald, 2009, When is a Liability not a Liability? Textual Analysis,
Dictionaries, and 10-Ks, Working Paper, University of Notre Dame.
Nagel, S., 2005, Short sales, institutional investors and the cross-section of stock returns, Journal
of Financial Economics 78, 277-309.
Rubinstein, Ariel, 1993, On Price Recognition and Computational Complexity in a Monopolistic
Model, Journal of Political Economy 101, 473-84.
Senchack, A.J., Jr., and L. Starks, 1993, Short-sale restrictions and market reaction to short-
interest announcements, Journal of Financial and Quantitative Analysis 28, 177-194.
Skoulakis, G., 2005, Assessment of Asset-Pricing Models using Cross-Sectional Regressions,
Working paper, Northwestern University.
Slovin, M. B., M. E. Sushka, and Y. M. Bendeck, 1991, The Intra-Industry Effects of Going-
Private Transactions, Journal of Finance 46, 1537-1550.
27
Tetlock, Paul C., 2007, Giving content to investor sentiment: The role of media in the stock
market, Journal of Finance 62, 1139-1168.
Tetlock, Paul C., 2008, All the News That’s Fit to Reprint: Do Investors React to Stale
Information?, Working Paper.
Tetlock, Paul C., 2009, Does Public Financial News Resolve Asymmetric Information?,
Working Paper.
Tetlock, Paul C., M. Saar-Tsechansky, and S. Macskassy, 2008, More Than Words: Quantifying
Language to Measure Firms' Fundamentals, Journal of Finance 63, 1437-1467.
Vega, C., 2006, Stock price reaction to public and private information, Journal of Financial
Economics 82, 103-133.
28
Figure 1
Volume around News Events
Figure 1 displays short volume, total volume, and the ratio of short volume to total volume for
the 15 days before and after news events. Short volume and total volume are scaled by their
mean values over the period -16 to -30. Panel A displays volume around all news events.
Panels B and C display volume for negative news events only. Version 1 (Panel B) uses a
negative sentiment variable that was constructed using the Harvard-IV-4 Dictionary as in
Tetlock (2007) and Engelberg (2008) while version 2 (Panel C) uses a sentiment measure that
employs the negative word list developed by Loughran and McDonald (2009).
Panel A: All News Events
Panel B: Negative News Events – version 1
Panel C: Negative News Events – version 2
0.0000
0.2000
0.4000
0.6000
0.8000
1.0000
1.2000
1.4000
1.6000
1.8000
-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10+11+12+13+14+15
Vo
lum
e
All News Events
Short Volume Total Volume Short Vol / Total Vol
0.0000
0.2000
0.4000
0.6000
0.8000
1.0000
1.2000
1.4000
1.6000
-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10+11+12+13+14+15
Vo
lum
e
Negative News Events - Version 1
Short Volume Total Volume Short Vol / Total Vol
0.0000
0.2000
0.4000
0.6000
0.8000
1.0000
1.2000
1.4000
1.6000
-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10+11+12+13+14+15
Vo
lum
e
Negative News Events - Version 2
Short Volume Total Volume Short Vol / Total Vol
29
Figure 2
Example Short Volume Portfolio Returns following News Events
Figure 2 displays buy and hold portfolio returns for a 12 month period following news events.
Each day for each news event, two portfolios are formed: the first portfolio consists of those
firms that had a specific news event and had low short volume as a percentage of total volume;
the second portfolio consists of those that had the news event and had high short volume as a
percentage of total volume. We then form control portfolios using a sample of firms that did not
experience a news event but were similar in terms of bid-ask-spread, institutional ownership,
market capitalization, and the number of news events over the previous month. The detailed
results are shown in Table VIII and three example results are shown below. Panel A displays
portfolio returns following dividend news and the returns for the matched control sample. Panel
B displays portfolio returns following earnings news and the associated control returns and Panel
C contains returns following news about insider stock sales and the associated control returns.
Panel A: Dividends
News Sample Control Sample
Panel B: Earnings
News Sample Control Sample
Panel C: Insider Stock Sales
News Sample Control Sample
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 1 2 3 4 5 6 7 8 9 10 11 12
Bu
y a
nd
Ho
ld R
etu
rn
Months after Portfolio Formation
Low Short Volume High Short Volume
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 1 2 3 4 5 6 7 8 9 10 11 12
Bu
y a
nd
Ho
ld R
etu
rn
Months after Portfolio Formation
Low Short Volume High Short Volume
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
0 1 2 3 4 5 6 7 8 9 10 11 12
Bu
y a
nd
Ho
ld R
etu
rn
Months after Portfolio Formation
Low Short Volume High Short Volume
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
0 1 2 3 4 5 6 7 8 9 10 11 12
Bu
y a
nd
Ho
ld R
etu
rn
Months after Portfolio Formation
Low Short Volume High Short Volume
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
0 1 2 3 4 5 6 7 8 9 10 11 12
Bu
y a
nd
Ho
ld R
etu
rn
Months after Portfolio Formation
Low Short Volume High Short Volume
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
0 1 2 3 4 5 6 7 8 9 10 11 12
Bu
y a
nd
Ho
ld R
etu
rn
Months after Portfolio Formation
Low Short Volume High Short Volume
30
Table I
Summary Statistics
The database has 1,888,868 observations over the period January 3, 2005 through July 6, 2007.
Panel A provides summary statistics at the firm level. News articles may be reissued throughout
the day as more information becomes available; in such situations we consider all of the related
article updates to be one unique news event and we keep track of the number of articles that are
rolled-into this unique news event. News Articles per Firm-Day is a count of all news articles
including reissued (updated) articles while Unique News Events per Firm-Day is a count of the
unique stories, excluding subsequent updates to an article. Short Vol. / Total Vol. is the short
volume from the NYSE TAQ Regulation SHO database as a percentage of total volume; Exempt
and Non-Exempt denote market maker short sales (exempt) from non-market maker short sales
(non-exempt), see section II.A for details. Market Capitalization is from CRSP. Panel B
contains summary statistics on the frequency of each news category in the database as well as
the mean number of negative words as a percentage of total words in the headline and body text
of each article. News articles may be classified into more than one category and we adopt two
methods for counting the number of negative words in the headline and body text of each article:
Version 1 uses the Harvard-IV-4 Dictionary as in Tetlock (2007) and Engelberg (2008) while
Version 2 uses the negative word list developed by Loughran and McDonald (2009).
Panel A – Firm Level Statistics Mean Median 1
st
Percentile
99
th
Percentile
Standard
Deviation
News Articles per Firm-Day 1.10 0.00 0.00 18.00 4.09
Unique News Events per Firm-Day 0.81 0.00 0.00 12.00 2.77
Short Vol. / Total Vol. 19.60% 17.47% 0.52% 62.46% 27.22%
Short Vol. / Total Vol. – Exempt 3.61% 1.44% 0.01% 32.26% 8.11%
Short Vol. / Total Vol. – Non-exempt 17.63% 15.66% 0.39% 55.37% 26.76%
Market Capitalization ($ mm) $5,856 $1,228 $32 $80,360 $19,329
Panel B – News Categories N
Mean Negative Headline
Words (% of total)
Mean Negative Body
Text Words (% of total)
Version 1 Version 2 Version 1 Version 2
10K 1,320 6.65% 4.07% 3.35% 1.79%
8K 10,803 4.83% 1.80% 2.91% 1.10%
Acquisitions, Mergers, Takeovers 56,993 5.09% 1.36% 3.18% 1.00%
Analysts' Comments & Ratings 49,508 5.00% 2.12% 3.16% 1.08%
Annual Meetings 4,041 6.63% 1.32% 3.01% 0.93%
Antitrust News 5,217 7.40% 3.88% 3.99% 1.72%
Bankruptcy-Related Filings 6,258 6.88% 3.30% 4.47% 1.99%
Bond Ratings & Comments 15,343 6.39% 1.62% 3.29% 1.15%
Buybacks 6,269 4.94% 1.03% 3.01% 0.93%
Contracts, Defense 4,734 6.03% 1.85% 3.24% 1.07%
Contracts, Government (not defense) 3,321 5.92% 1.44% 3.20% 1.01%
31
Table I (continued)
Panel B – News Categories N
Mean Negative Headline
Words (% of total)
Mean Negative Body
Text Words (% of total)
Version 1 Version 2 Version 1 Version 2
Contracts, Nongovernment 19,102 5.96% 1.16% 2.93% 0.82%
Corporate Governance 4,981 7.00% 2.49% 3.96% 1.55%
Corporate Restructurings 5,631 6.04% 2.16% 3.78% 1.43%
Divestitures or Asset Sales 11,587 4.85% 1.25% 3.21% 1.03%
Dividend News 24,731 7.64% 0.60% 2.64% 0.54%
Earnings 40,705 5.84% 1.09% 3.12% 0.89%
Earnings Projections 37,432 5.57% 1.39% 3.26% 0.99%
Financing Agreements 6,919 5.01% 1.27% 2.95% 0.98%
High-Yield Issuers 173,357 5.24% 1.68% 2.92% 0.90%
Initial Public Offerings 9,351 3.70% 1.01% 2.59% 0.76%
Insider Stock Buys 21,489 1.65% 0.45% 1.15% 0.40%
Insider Stock Sells 54,868 1.68% 1.29% 1.34% 0.34%
Joint Ventures 9,081 5.71% 1.20% 3.03% 0.88%
Labor Issues 12,376 7.12% 2.49% 3.89% 1.44%
Lawsuits 15,351 7.23% 3.85% 4.37% 2.40%
Leveraged Buyouts 2,286 6.07% 1.46% 3.79% 1.16%
Management Issues 13,840 5.58% 1.71% 3.38% 1.18%
Market News 14,068 6.16% 2.01% 3.77% 1.31%
Money Market News 1,298 6.61% 1.99% 4.00% 1.37%
New Products & Services 24,583 7.08% 1.24% 3.02% 0.77%
Personnel Appointments 29,994 6.71% 1.36% 3.19% 0.86%
Point of View 17,316 6.09% 1.84% 3.88% 1.26%
Product Distribution 2,440 5.90% 1.52% 3.21% 0.99%
Research & Development 5,323 6.80% 1.88% 3.76% 1.21%
Spinoffs 1,874 3.89% 1.17% 2.68% 0.85%
Stock Options 5,679 5.42% 1.95% 3.38% 1.27%
Stock Ownership 25,567 2.69% 0.70% 2.38% 0.34%
Stock Splits 2,230 4.23% 0.56% 2.13% 0.35%
32
Table II
Regression Analysis of Short Volume Ratio around News Events
Table II contains the results of six regressions of short sales volume on a set of indicator variables representing stories in each of the
news categories. Specifically, the dependent variable is aggregate short volume as a percentage of total volume and the independent
variables are indicator variables that take the value one if there is a news story in a particular news category and zero otherwise. We
vary the timing of the dependent variable relative to the news event in order to examine short volume changes around news. For
example, t-2 indicates that the dependent variable is observed two days prior to the news event. For After Minus Before the dependent
variable is the difference in the short volume ratio between dates t+2 and t-2. To control for the documented response of short sellers
to past returns, we include two lags of daily returns. *** indicates significance at the 1% level, ** indicates significance at the 5%
level, and * indicates significance at the 10% level.
After
Event Time of the Dependent Variable Minus
News Events t-2 t-1 t=0 t+1 t+2 Before
Mean of the fixed effects 0.1823*** 0.1855*** 0.1844*** 0.1835*** 0.185*** -0.0102
Return (1 day lag) 0.3878*** 0.3854*** 0.3882*** 0.3893*** 0.3904*** 0.5749***
Return (2 day lag) 0.2715*** 0.2729*** 0.2735*** 0.2738*** 0.2744*** -0.4506***
10K -0.0007 -0.0036 -0.0039 -0.0031 -0.0018 0.0046
8K -0.0011 -0.0019 -0.0003 -0.0003 -0.0003 0.0012
Acquisitions, Mergers, Takeovers 0.0023 0.0006 0.0001 0.0000 0.0006 0.0268
Analysts' Comments & Ratings of Stocks 0.0017 0.0023 0.0123*** 0.006*** 0.0063*** 0.0078**
Annual Meetings -0.0084** -0.0022 -0.0016 -0.0037 -0.0013 -0.0016
Antitrust News -0.0041 -0.0059 -0.0028 -0.0046 -0.0053 0.0005
Bankruptcy-Related Filings 0.0035 0.0011 -0.0004 -0.0035 -0.0013 -0.0124
Bond Ratings & Comments 0.0005 -0.0005 -0.0005 -0.0018 -0.0012 0.0801
Buybacks -0.0038 -0.0052* -0.0017 -0.0041 -0.003 0.0018
Contracts, Defense -0.0021 -0.0028 0.0078 0.0025 0.0009 0.0092
Contracts, Government (not defense) -0.0059 -0.0084* -0.0039 -0.0010 -0.0015 0.0098
Contracts, Nongovernment 0.0008 0.0014 -0.0012 0.0003 -0.0006 -0.0024
Corporate Governance -0.0008 -0.0026 -0.0004 -0.0001 0.0008 0.0037
Corporate Restructurings 0.0033 0.0021 0.0004 -0.0024 -0.0001 0.0044
Divestitures or Asset Sales -0.0006 -0.0011 -0.0010 -0.0018 0.0000 0.0005
Dividend News 0.0015 -0.0034** -0.0005 0.0003 0.0005 0.0048*
Earnings 0.0008 -0.0027* -0.0025 0.0028* 0.0049*** 0.0132***
33
Table II (continued) After
Event Time of the Dependent Variable Minus
News Events t-2 t-1 t=0 t+1 t+2 Before
Earnings Projections -0.0036* 0.0004 -0.0007 0.0031 0.0025 0.007**
Financing Agreements -0.0017 -0.0035 -0.0038 -0.0006 -0.0030 0.0013
High-Yield Issuers -0.0004 0.0019 0.0057** 0.0035 0.0024 0.0046
Initial Public Offerings -0.0030 0.0015 0.0009 -0.0025 -0.0009 -0.0038
Insider Stock Buys -0.0024 -0.0029 -0.0021 -0.0014 -0.0023 -0.0002
Insider Stock Sells -0.0058** -0.0059** -0.0066** -0.0040 -0.0011 0.0069
Joint Ventures -0.0046 -0.0051* -0.0017 0.0028 -0.0004 0.0123**
Labor Issues -0.0013 -0.0006 -0.0016 0.0009 -0.0008 0.0033
Lawsuits 0.0053* -0.0014 -0.0053* -0.0003 0.0011 0.0085
Leveraged Buyouts -0.0031 0.0145*** -0.0119** -0.0025 -0.0063 -0.0233**
Management Issues -0.0036 -0.0024 -0.0047* 0.0000 -0.0020 -0.0048
Market News -0.0030 -0.0037 -0.0057* -0.0042 -0.0056* -0.0034
Money Market News -0.0018 -0.0019 -0.0105 0.0006 -0.0083 0.0025
New Products & Services 0.0021 -0.0025 -0.0020 0.0001 0.0017 0.0013
Personnel Appointments -0.0012 0.0024 -0.0002 0.0013 0.0034* 0.0043
Point of View -0.0011 0.0016 -0.0038 -0.0028 -0.002 0.0025
Product Distribution -0.0054 -0.0035 -0.0042 -0.0012 -0.0028 0.0055**
Research & Development -0.0019 -0.0003 -0.0058 -0.0051 -0.0021 -0.0010
Spinoffs -0.0057 -0.0037 -0.0069 -0.0059 -0.0087 -0.0044
Stock Options -0.0030 -0.0031 -0.0040 -0.0038 -0.0035 -0.0076
Stock Ownership -0.0045*** -0.0042*** -0.0041*** -0.0023* -0.0010 0.0041
Stock Splits 0.0017 0.0000 0.0047 0.0018 0.0102*** -0.0032
34
Table III
Regression Analysis of Raw Short Volume around News Events
Table III contains the results of six regressions of short sales volume on a set of indicator variables representing stories in each of the
news categories. Specifically, the dependent variable is the raw aggregate short volume (not scaled) and the independent variables are
indicator variables that take the value one if there is a news story in a particular news category and zero otherwise. We vary the
timing of the dependent variable relative to the news event in order to examine short volume changes around news. For example, t-2
indicates that the dependent variable is observed two days prior to the news event. For After Minus Before the dependent variable is
the difference in the short volume ratio between dates t+2 and t-2. To control for the documented response of short sellers to past
returns, we include two lags of daily returns. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and
* indicates significance at the 10% level.
After
Event Time of the Dependent Variable Minus
News Events t-2 t-1 t=0 t+1 t+2 Before
Mean of the fixed effects 141,452 137,444 128,403 120,962* 144,973 -123,750***
Return (1 day lag) 351,936*** 353,335*** 335,762*** 331,588*** 342,662*** 549,513***
Return (2 day lag) 160,030*** 160,302*** 157,685*** 152,383*** 151,151*** -761,974***
10K -23,762*** -28,923*** 23,824*** 24,974*** 1,230 11,559
8K 7,775** 7,171* 14,523*** 14,654*** 3,666 72,636***
Acquisitions, Mergers, Takeovers 1,946 -1,256 11,297*** 8,393** 6,162* -119,140**
Analysts' Comments & Ratings of Stocks 4,658 31,158*** 77,550*** 26,301*** 11,636*** 5,678
Annual Meetings 3,465 15,713*** 20,810*** 12,863** 8,636 -20,271
Antitrust News 3,173 -7,773 26,145*** 3,301 10,656* 65,438***
Bankruptcy-Related Filings 15,192** 22,822*** 90,670*** 42,514*** 42,438*** 47,730***
Bond Ratings & Comments 7,069* 29,597*** 77,503*** 28,790*** 5,213 -141,097
Buybacks 19,459*** 21,194*** 92,991*** 53,651*** 19,574*** 33,025***
Contracts, Defense -31,532*** -27,023** -11,273 -10,146 -10,117 54,981**
Contracts, Government (not defense) -13,446 -18,100** -3,401 1,168 -3,031 27,580
Contracts, Nongovernment 4,944 2,579 14,985*** 12,979*** 4,351 8,176
Corporate Governance 20,671*** 29,455*** 44,600*** 27,269*** 13,420** 2,389
Corporate Restructurings 806 10,367* 57,111*** 37,131*** -7,468 7,584
Divestitures or Asset Sales 4,527 9,350** 38,324*** 10,396** 2,118 -3,634
Dividend News 6,770*** 2,951 12,491*** 17,614*** 8,102*** 16,521***
Earnings -8,809*** 1,668 63,417*** 65,849*** 25,914*** 101,013***
35
Table III (continued)
After
Event Time of the Dependent Variable Minus
News Events t-2 t-1 t=0 t+1 t+2 Before
Earnings Projections -4,336 3,161 53,246*** 44,373*** 16,240*** 63,330***
Financing Agreements 6,598 2,358 29,615*** 13,609*** -147 2,576
High-Yield Issuers -3,626 -2,538 37,108*** 4,242 1,661 11,645
Initial Public Offerings 17,781*** 53,697*** 60,989*** 19,461*** 9,143 -36,536**
Insider Stock Buys 14,979*** 17,307*** 8,147** 4,510 1,752 -26,275***
Insider Stock Sells 18,020*** 5,991 -1,482 4,197 6,098 -10,966
Joint Ventures 10,298** -528 8,540* 10,657** 4,212 9,445
Labor Issues 6,204 14,395*** 65,368*** 32,374*** 22,333*** 37,023***
Lawsuits 4,035 14,461*** 16,622*** 15,239*** 1,014 79,158
Leveraged Buyouts 25,052*** 56,619*** 172,693*** 107,072*** 46,789*** 55,578***
Management Issues -4,362 1,217 -6,665 -820 -4,285 85,346***
Market News 6,757 22,521*** 199,417*** 87,374*** 22,841*** -551
Money Market News 89,479*** 135,472*** 291,721*** 133,469*** 64,114*** -172
New Products & Services -2,304 -6,230 -3,791 2,751 3,359 65,542***
Personnel Appointments 2,044 -711 13,598*** 6,932** 3,177 -6,578
Point of View 19,505*** 24,110*** 30,348*** 15,518** 12,959* 13,839
Product Distribution -14,420** 6,252 6,156 6,571 -17,414** 4,708
Research & Development 21,134** -6,816 3,107 -9,898 -3,898 -8,813
Spinoffs -1,126 77,486*** 66,522*** 42,935*** 29,173*** -7,178
Stock Options 6,848 17,998*** 97,709*** 47,563*** 42,051*** 16,458
Stock Ownership 2,128 7,942*** 9,474*** 8,505*** 3,436 7,098
Stock Splits -4,078 27,491*** 14,689** 433 -3,889 -29,274
36
Table IV
Cross-Sectional Relation between Returns, Short Sales, and News
Table IV contains the results from Fama-MacBeth (1973) type regressions using daily
observations over the period January 3, 2005 through July 6, 2007. The regressions are done
firm by firm, and the dependent variable is the buy and hold (compound) return over the next 20
trading days. Panel A is calculated using raw returns as the dependent variable while Panel B
uses characteristic adjusted returns as in Daniel, Grinblatt, Titman, and Wermers (1997),
however we omit the book to market factor due to missing Compustat data for some firms. The
Short Volume Ratio is daily short volume / total volume. News Event is an indicator variable
that takes the value one if a news event occurs for a particular stock, and Short-News Interaction
is the product of Short Volume Ratio and the News Event indicator. Returnt=0 is the return on
each stock on the day that short volume and news are observed. T-statistics are below the
parameter estimates in italics and are calculated using Newest-West (1987) standard errors with
20 lags. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and
* indicates significance at the 10% level.
Panel A: Raw Returns
Model
(1) (2) (3) (4) (5)
Intercept 0.0171*** 0.0173*** 0.0171*** 0.0169*** 0.0168***
(4.70) (4.78) (4.73) (4.82) (4.99)
Short Volume Ratio -0.0053** -0.0053** -0.0044* -0.0047** -0.0052**
(-2.19) (-2.20) (-1.86) (-2.01) (-2.24)
News Event -0.0010 0.0000 0.0000 0.0000
(-1.37) (0.02) (0.05) (0.06)
Short – News Interaction -0.0050** -0.0053*** -0.0053***
(-2.41) (-2.62) (-2.72)
Returnt=0 0.0213 0.0227
(1.29) (1.35)
Returnt=-1 0.0287*
(1.86)
Returnt=-2 0.0356**
(2.38)
Panel B: DGTW Returns
Intercept 0.0054*** 0.0056*** 0.0054*** 0.0052*** 0.0051***
(5.32) (5.16) (4.97) (4.62) (4.07)
Short Volume Ratio -0.0069*** -0.0068*** -0.0058*** -0.0061*** -0.0065***
(-3.27) (-3.27) (-2.85) (-3.00) (-3.23)
News Event -0.0009* 0.0003 0.0003 0.0003
(-1.72) (0.35) (0.42) (0.44)
Short – News Interaction -0.0059*** -0.0063*** -0.0063***
(-2.75) (-2.98) (-3.12)
Returnt=0 0.0216 0.0226
(1.37) (1.41)
Returnt=-1 0.0303**
(2.03)
Returnt=-2 0.0333**
(2.33)
37
Table V
Cross-Sectional Relation between Returns, Short Sales, and News for Non-Exempt Trades
Table V contains the results from Fama-MacBeth (1973) type regressions using daily
observations over the period January 3, 2005 through July 6, 2007. The sample only includes
those short sales transactions that were classified as non-exempt as discussed in Section II.A of
the text. The regressions are done firm by firm, and the dependent variable is the buy and hold
(compound) return over the next 20 trading days. Panel A is calculated using raw returns as the
dependent variable while Panel B uses characteristic adjusted returns as in Daniel, Grinblatt,
Titman, and Wermers (1997), however we omit the book to market factor due to missing
Compustat data for some firms. The Short Volume Ratio is daily short volume / total volume.
News Event is an indicator variable that takes the value one if a news event occurs for a
particular stock, and Short-News Interaction is the product of Short Volume Ratio and the News
Event indicator. Returnt=0 is the return on each stock on the day that short volume and news are
observed. T-statistics are below the parameter estimates in italics and are calculated using
Newest-West (1987) standard errors with 20 lags. *** indicates significance at the 1% level, **
indicates significance at the 5% level, and * indicates significance at the 10% level.
Panel A: Raw Returns
Model
(1) (2) (3) (4) (5)
Intercept 0.0179*** 0.0181*** 0.0179*** 0.0177*** 0.0176***
(4.87) (4.97) (4.92) (5.02) (5.16)
Short Volume Ratio -0.0077*** -0.0076*** -0.0067*** -0.0072*** -0.0075***
(-2.98) (-2.98) (-2.67) (-2.84) (-2.94)
News Event -0.0012* -0.0002 -0.0002 -0.0002
(-1.48) (-0.26) (-0.20) (-0.17)
Short – News Interaction -0.0049* -0.0055** -0.0056**
(-1.95) (-2.25) (-2.33)
Returnt=0 0.0395** 0.0366**
(2.16) (2.12)
Returnt=-1 0.0236**
(2.34)
Returnt=-2 0.0221
(1.54)
Panel B: DGTW Returns
Intercept 0.0061*** 0.0063*** 0.0061*** 0.0060*** 0.0059***
(5.45) (5.27) (5.08) '(4.80) (4.56)
Short Volume Ratio -0.0092*** -0.0091*** -0.0081*** -0.0085*** -0.0088***
(-3.85) (-3.84) (-3.44) '(-3.59) (-3.66)
News Event -0.0011** 0.0001 0.0002 0.0002
(-1.53) (0.11) '(0.21) (0.23)
Short – News Interaction -0.0063** -0.0069*** -0.0070***
(-2.51) '(-2.84) (-2.94)
Returnt=0 0.0390** 0.0351**
'(2.23) (2.10)
Returnt=-1 0.0258***
(2.70)
Returnt=-2 0.0223
(1.65)
38
Table VI
Cross-Sectional Relation between Returns, Short Sales, and News for Exempt Trades
Table VI contains the results from Fama-MacBeth (1973) type regressions using daily
observations over the period January 3, 2005 through July 6, 2007. The sample only includes
those short sales transactions that were classified as exempt as discussed in Section II.A of the
text. The regressions are done firm by firm, and the dependent variable is the buy and hold
(compound) return over the next 20 trading days. Panel A is calculated using raw returns as the
dependent variable while Panel B uses characteristic adjusted returns as in Daniel, Grinblatt,
Titman, and Wermers (1997), however we omit the book to market factor due to missing
Compustat data for some firms. The Short Volume Ratio is daily short volume / total volume.
News Event is an indicator variable that takes the value one if a news event occurs for a
particular stock, and Short-News Interaction is the product of Short Volume Ratio and the News
Event indicator. Returnt=0 is the return on each stock on the day that short volume and news are
observed. T-statistics are below the parameter estimates in italics and are calculated using
Newest-West (1987) standard errors with 20 lags. *** indicates significance at the 1% level, **
indicates significance at the 5% level, and * indicates significance at the 10% level.
Panel A: Raw Returns
Model
(1) (2) (3) (4) (5)
Intercept 0.0149*** 0.0148*** 0.0147*** 0.0140*** 0.0138***
(4.32) (4.31) (4.32) (4.26) (4.34)
Short Volume Ratio 0.0202** 0.0196*** 0.0203*** 0.0200*** 0.0191***
(5.14) (4.89) (4.77) (4.48) (4.13)
News Event 0.0039* 0.0025 0.0006 -0.0051
(1.90) (0.14) (0.04) (-0.20)
Short – News Interaction 0.2099 0.2189 0.3374
(0.75) (0.83) (0.84)
Returnt=0 -0.0845** -0.1198***
(-2.26) (-3.03)
Returnt=-1 0.1459***
(2.91)
Returnt=-2 -0.1224***
(-3.20)
Panel B: DGTW Returns
Intercept 0.0033 0.0032 0.0031 0.0025 0.0022
(1.53) (1.49) (1.47) (1.19) (1.09)
Short Volume Ratio 0.0204*** 0.0200*** 0.0205*** 0.0201*** 0.0191***
(5.75) (5.55) (5.54) (5.15) (4.72)
News Event 0.0030 0.0048 0.0028 -0.0017
(1.62) (0.29) (0.17) (-0.08)
Short – News Interaction 0.158 0.1806 0.2748
(0.64) (0.73) (0.77)
Returnt=0 -0.0815** -0.1198***
(-2.25) (-3.20)
Returnt=-1 0.1602***
(3.58)
Returnt=-2 -0.1407***
(-3.67)
39
Table VII
Equity Returns Following Specific News Events
Table VII examines equity returns following news events according to the model:
where the dependent variable is the compound excess return from day 1 to day 20 following the news event, ret0 is the excess return
on the day of the news event, and Size is measured using the market capitalization for each firm. Regressions are run individually for
each news event and only when a news event occurs. Firm fixed effects are included and the intercept is the average of the fixed
effects. T-statistics are reported below. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and *
indicates significant at the 10% level. Intercept Return (t=0) Short Volume Size
News Events Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat
10K 0.0119 0.03 -0.5421 -2.31** 0.0675 0.57 -0.0024 -1.17
8K 0.0223 0.33 -0.1218 -2.20** -0.0079 -0.45 -0.0025 -4.92***
Acquisitions, Mergers, Takeovers 0.0383 0.19 -0.0333 -0.61 -0.0194 -1.26 -0.0067 -5.48***
Analysts' Comments & Ratings 0.0372 0.86 0.0499 1.74* -0.0232 -1.90* -0.0077 -5.89***
Annual Meetings 0.0188 0.96 -0.3073 -1.93* 0.0037 0.14 -0.0012 -2.29**
Antitrust News 0.0118 -0.38 0.3671 3.34*** 0.0051 0.20 -0.0007 -1.42
Bankruptcy-Related Filings 0.0144 -1.55 -0.1910 -1.73* 0.0323 0.86 -0.0016 -2.60***
Bond Ratings & Comments 0.0189 0.37 -0.0117 -0.28 -0.0168 -1.11 -0.0012 -2.50**
Buybacks 0.0186 -0.17 -0.0337 -0.55 -0.0060 -0.29 -0.0009 -2.39**
Contracts, Defense 0.0080 0.33 -0.2923 -1.01 -0.0012 -0.02 -0.0007 -1.08
Contracts, Government (not defense) 0.0247 0.21 -0.0249 -0.11 -0.0300 -0.72 -0.0008 -0.95
Contracts, Nongovernment 0.0323 0.35 -0.1014 -1.24 -0.0229 -1.20 -0.0035 -4.39***
Corporate Governance 0.0416 -1.27 -0.4701 -3.05*** -0.0207 -0.61 -0.0027 -3.64***
Corporate Restructurings 0.0355 0.07 -0.0316 -0.44 -0.0953 -3.24*** -0.0009 -1.65
Divestitures or Asset Sales 0.0325 0.40 0.1591 2.24** -0.0467 -2.29** -0.0022 -3.49***
Dividend News 0.0272 2.04** -0.1339 -4.02*** -0.0047 -0.61 -0.0033 -7.01***
Earnings 0.0412 0.36 0.1145 5.25*** -0.0271 -3.08*** -0.0080 -9.85***
Earnings Projections 0.0445 0.07 0.0594 2.28** -0.0435 -3.76*** -0.0073 -7.48***
Financing Agreements 0.0215 0.58 -0.1209 -1.37 -0.0213 -0.88 -0.0016 -2.81***
High-Yield Issuers 0.0358 1.40 0.0226 0.67 -0.0137 -1.09 -0.0116 -5.68***
Initial Public Offerings 0.0107 -1.41 0.4056 2.91*** 0.0592 2.16** -0.0013 -2.05**
40
Table VII (continued)
Intercept Return (t=0) Short Volume Size
News Events Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat
Insider Stock Buys 0.0194 0.13 -0.2852 -3.03*** -0.0012 -0.08 -0.0017 -3.57***
Insider Stock Sells 0.0279 -1.49 -0.2876 -2.18** -0.0208 -0.90 -0.0043 -2.94***
Joint Ventures 0.0318 1.06 0.0844 0.69 -0.0189 -0.75 -0.0028 -3.77***
Labor Issues 0.0372 0.32 0.0647 1.07 -0.0267 -1.25 -0.0034 -4.16***
Lawsuits 0.0417 0.50 -0.0004 0.00 -0.0442 -1.78* -0.0037 -3.78***
Leveraged Buyouts 0.0070 -0.25 -0.2145 -1.26 -0.0018 -0.04 -0.0003 -0.43
Management Issues 0.0382 0.25 0.1289 1.62 -0.0439 -2.06** -0.0040 -5.12***
Market News 0.0416 0.08 -0.0841 -1.68* 0.0054 0.22 -0.0040 -5.36***
Money Market News 0.0375 1.16 -0.6654 -2.71*** -0.0742 -1.22 -0.0006 -1.54
New Products & Services 0.0340 -0.01 -0.1535 -1.31 -0.0819 -4.13*** -0.0021 -2.36**
Personnel Appointments 0.0325 0.22 -0.0694 -1.19 -0.0170 -1.24 -0.0052 -6.20***
Point of View 0.0294 1.55 0.0633 0.43 0.0251 0.65 -0.0040 -2.53**
Product Distribution 0.0260 -1.43 -0.3118 -1.87* -0.0397 -1.11 -0.0013 -2.16**
Research & Development 0.0310 2.34** -0.2311 -1.25 -0.0964 -2.35** -0.0008 -1.31
Spinoffs 0.0315 -0.59 -0.2707 -1.67* -0.0325 -0.40 -0.0011 -1.34
Stock Options 0.0178 0.26 0.0024 0.03 0.0144 0.37 -0.0015 -2.47**
Stock Ownership 0.0318 0.14 -0.0153 -0.35 -0.0225 -3.25*** -0.0042 -8.09***
Stock Splits 0.0234 -0.95 -0.0146 -0.13 -0.0410 -1.74* -0.0006 -1.04
Fisher Stat 160.37***
Fisher P-Value 0.00%
41
Table VIII
Short Volume Portfolio Returns following News Events
Table VIII displays buy and hold portfolio returns for a 12 month period following news events. Each day for each news event, two
portfolios are formed: the first portfolio consists of those firms that had a specific news event and had low short volume as a
percentage of total volume; the second portfolio consists of those that had the news event and had high short volume as a percentage
of total volume. In addition, we form control portfolios using a sample of firms that did not experience the news event but were
similar in terms of bid-ask-spread, institutional ownership, market capitalization, and the number of news events over the previous
month. Difference is the return of the High portfolio less the Low portfolio, and Difference in Difference is the Difference value of the
Control Sample less the Difference value of the Event Sample. *** indicates significance at the 1% level, ** indicates significance at
the 5% level, and * indicates significant at the 10% level. Event Sample: 12 Month Returns Control Sample: 12 Month Returns Difference in
News Events Low High Difference Low High Difference Difference
10K 3.29% 2.23% -1.06% 3.07% 3.99% 0.91% 1.98%
8K 3.56% -1.25% -4.81% 5.52% 3.18% -2.35% 2.46%
Acquisitions, Mergers, Takeovers 4.59% 3.59% -0.99% 4.45% 4.97% 0.52% 1.51%
Analysts' Comments & Ratings 4.61% -0.51% -5.12% 2.33% 4.26% 1.93% 7.05%
Annual Meetings 1.62% -1.25% -2.87% 5.43% 5.57% 0.14% 3.01%
Antitrust News 2.70% 0.81% -1.90% 9.31% 7.51% -1.80% 0.09%
Bankruptcy-Related Filings 5.52% 6.84% 1.32% 8.42% 1.18% -7.24% -8.56%
Bond Ratings & Comments 3.05% 2.17% -0.89% 2.30% 7.29% 4.99% 5.88%
Buybacks 3.77% 0.75% -3.02% 5.32% 3.16% -2.16% 0.86%
Defense Contracts 8.83% 4.37% -4.47% 3.43% 9.59% 6.15% 10.62%
Contracts, Defense 3.25% 5.41% 2.16% 5.68% -2.73% -8.41% -10.57%
Contracts Government (not defense) 3.27% 2.49% -0.78% 3.70% 5.63% 1.92% 2.71%
Corporate Governance 4.33% 2.54% -1.79% 1.21% 9.48% 8.27% 10.06%
Corporate Restructurings 4.19% 4.22% 0.03% 0.34% 9.86% 9.52% 9.49%
Divestitures or Asset Sales 2.15% 2.18% 0.03% 2.27% 6.88% 4.61% 4.58%
Dividend News 4.06% -0.33% -4.39% 3.55% 2.90% -0.65% 3.74%
Earnings 6.47% 1.02% -5.45% 5.84% 2.18% -3.66% 1.79%
Earnings Projections 4.64% 1.33% -3.31% 4.89% 4.02% -0.86% 2.45%
Financing Agreements 2.59% 2.04% -0.54% 4.23% 5.62% 1.40% 1.94%
High-Yield Issuers 3.25% -0.07% -3.32% 6.04% 7.14% 1.10% 4.42%
Initial Public Offerings 5.41% 6.55% 1.14% 5.68% 7.80% 2.11% 0.97%
Insider Stock Buys 0.25% -0.73% -0.98% 5.29% 4.54% -0.76% 0.23%
Insider Stock Sells 1.52% -1.76% -3.29% 4.59% 1.69% -2.90% 0.39%
42
Table VIII (continued)
Event Sample: 12 Month Returns Control Sample: 12 Month Returns Difference in
News Events Low High Difference Low High Difference Difference
Joint Ventures 4.59% 3.62% -0.97% 3.70% 5.59% 1.89% 2.86%
Labor Issues 2.09% 2.91% 0.82% 5.60% 6.45% 0.86% 0.04%
Lawsuits 7.01% 3.50% -3.51% 4.65% 6.34% 1.70% 5.21%
Leveraged Buyouts 1.97% 2.24% 0.27% 5.75% 7.88% 2.14% 1.87%
Management Issues 3.90% -0.63% -4.53% 3.99% 5.30% 1.31% 5.84%
Market News 3.49% 0.95% -2.54% 6.37% 4.44% -1.93% 0.61%
Money Market News 5.73% 7.34% 1.61% 3.68% 19.59% 15.91% 14.29%
New Products & Services 5.22% -0.17% -5.38% 7.95% 4.99% -2.97% 2.42%
Personnel Appointments 5.98% 0.64% -5.33% 3.32% 3.36% 0.03% 5.37%
Point of View 8.11% 0.57% -7.54% 4.09% 4.91% 0.83% 8.36%
Product Distribution 4.03% -2.46% -6.50% 0.67% 6.41% 5.74% 12.24%
Research & Development 3.69% 1.72% -1.98% 7.91% 5.19% -2.71% -0.74%
Spinoffs 1.80% 2.96% 1.15% 4.78% 2.74% -2.04% -3.19%
Stock Options 7.49% 5.31% -2.18% 7.76% 4.43% -3.33% -1.15%
Stock Ownership 2.68% 0.20% -2.48% 5.65% 3.72% -1.94% 0.54%
Stock Splits 4.66% 3.18% -1.48% 2.40% 2.04% -0.35% 1.12%
Mean 2.89%
Median 2.42%
T-statistic 3.73***
Wilcoxon Z Score 3.91***